For a single-antenna radio communications system, the existing modulation and coding technologies may approach the theoretical Shannon limit very well. Therefore, a multiple-antenna technology is developed to further increase the baud rate and the reliability of the radio communications system.
A conventional method for increasing the baud rate and the reliability by using the multiple-antenna technology is a space division multiplexing method based on the multiple input multiple output (Multiple Input Multiple Output, MIMO for short) technology. With the MIMO-based space division multiplexing method, the space division multiplexing is implemented mainly by introducing multiple antennas to both a user equipment (User Equipment, UE for short) and a base station device, for example, an eNodeB. If both a receiving end and a transmitting end have already known information about a channel matrix, an optimal MIMO system is capable of obtaining a high degree-of-freedom gain and a large system capacity by using a precoding scheme based on singular value decomposition implemented on the channel matrix. However, for the MIMO-based space division multiplexing method, computational complexity is high when the UE or the eNodeB estimates the channel matrix according to an orthogonal reference signal of each antenna, and because of the time-varying characteristic of the channel matrix, the UE or the eNodeB needs to continuously update the channel matrix estimation and provide feedback to a peer end, thereby causing a high overhead.
A beamforming method based on an array antenna is developed to solve the preceding problem. That is, mainly, the eNodeB implements beamforming on a single array antenna, then estimates a direction of arrival (Direction of Arrival, DoA for short) of a received signal in a beam direction to obtain the channel matrix, and implements precoding after the singular value decomposition according to the channel matrix. Compared with the conventional MIMO-based space division multiplexing method, when the beamforming method based on the array antenna is used, the UE is not required to implement channel estimation and uplink feedback, and the speed of updating the channel matrix is greatly reduced. However, because no degree-of-freedom gain is provided, the capability to increase the system capacity is far weaker than the MIMO-based space division multiplexing method. If the beamforming method based on the array antenna is used, multiple independent array antennas must be used to obtain the degree-of-freedom gain, thereby increasing the system complexity.
Therefore, how to obtain the degree-of-freedom gain becomes a problem to be solved urgently when a MIMO system is implemented based on the single array antenna.